Conventional communications networks require procedures and products that dynamically establish and maintain connections between devices attached to the network through signaling. It is important for such signaling to be, among other things, flexible and responsive to the needs of the network and the customers of that network. Addressing plays an important role in signaling, and provides the means for structure and order in conventional networks and the signaling that is carried out therein.
Asynchronous Transfer Mode (“ATM”) is an increasingly popular standard for high-speed communication. An information stream, whether it be data, voice, video or other type of information, is divided into packets called “ATM cells.” Each ATM cell is fifty-three (53) bytes in length. An ATM cell comprises two main sections, a header, which is five bytes in length, and a payload, which is forty-eight bytes in length. The payload includes or corresponds to at least part of the subject information stream. The header includes information corresponding to a path to a desired destination, or endpoint, for the cell.
An ATM System typically comprises three architecture layers. An “Adaptation layer” divides information that it receives, whether it be data, voice, video or other type of information, into one or more (as needed) forty-eight byte payloads. An “ATM layer” adds a five-byte header comprising addressing information to each forty-eight byte payload. Once joined together, the five-byte header and the forty-eight byte payload comprise an ATM cell. A “Physical layer” converts the ATM cell to appropriate electrical, optical, or other format for physical transport.
The header comprises a virtual path identifier (VPI) and a virtual channel identifier (VCI). Within a typical ATM system, virtual connections are established between system elements as needed according to the VPI and VCI contained in the header. The header provides information which facilitates virtual connections between network elements.
For an introduction to ATM, see David E. McDysan & Darren L. Spohn, ATM Theory and Application (McGraw-Hill, Inc. 1995), the disclosure of which is incorporated herein by reference. For a further introduction to ATM and a description of various standards and specifications related to ATM, see The ATM Forum Technical Committee, User-Network Interface (UNI) Specification Version 3.1 (1994), the disclosure of which is incorporated herein by reference.
In a conventional ATM network, each ATM connected endpoint, or point of attachment, which can be a device such as a telephone, computer, or video monitor, for instance, has an address. In one embodiment of such a network, when a first ATM device wishes to establish a connection with a second ATM device, the first ATM device sends a SETUP message to the ATM switch connected to it (“first ATM switch”). This message includes addressing information, in digital form, including the ATM address of the second ATM device. This first ATM switch examines the SETUP message. In particular, this first ATM switch examines the included address of the second ATM device. This first switch determines which switch in the network the SETUP message should be sent to next (assuming that the first switch is not directly connected to the second ATM device) and forwards the message to a second switch. Similarly, the second switch examines the SETUP message and determines which switch in the network the message should be sent to next (same assumption) and forwards the packet to a third switch. This process continues until the SETUP message arrives at the second ATM device (or “endpoint”).
When the SETUP message arrives at the second ATM device, and if the device can support the desired connection, the second ATM device returns a CONNECT message to the first ATM device. As the CONNECT messages returns through the network switches back to the first ATM device, the switches set up a virtual connection, or virtual circuit, between the first ATM device and the second ATM device. In a conventional network, a CONNECT message includes the VPI/VCI values that the first ATM device should use for ATM cells that it wishes to send to the second ATM device. These VPIVCI values are integrated into the ATM cells at the first ATM device.
Several ATM address formats have been developed. The references cited above describe these AESAs in detail. Conventional public addresses are based upon the ITU-T E.164 format (or “native E.164” format). This format is generally an Integrated Services Digital Network (ISDN) telephone number. For example, a native E.164 address for a telephone in the Atlanta, Ga. area might be 14045551212. This number, by its 404 numbering plan area (or “NPA”) designation, is a geographic address indicating the Atlanta, Ga. area. The native E.164 address is based on the geographical location of the user. The digits of such an address generally includes the area code and, for international calls, the country code. The length of a native E.164 address is variable, depending upon, for example, whether the call made is an international call.
Conventional private ATM addresses are known as ATM End System Addresses (AESAs). AESAs are fixed in length at twenty (20) bytes. The ATM Forum supports at least three conventional AESA formats: E.164 AESA, Data Country Code (or “DCC”) AESA, and International Code Designator (or “ICD”) AESA. These formats are discussed herein as they relate to, and are used in, the United States.
In all three formats, the first thirteen (13) bytes are called the “network prefix” and the second seven (7) bytes are called the “user part.” In all three formats, the first byte of the network prefix (also the first byte of the AESA) is used for an authority and format identifier (or “AFI”). The AFI identifies which addressing scheme is found in the subsequent nineteen bytes. The E.164 AESA is identified by an AFI value of 45 (hex), the DCC AESA is identified by an AFI value of 39 (hex), and the ICD AESA is identified by an AFI value of 47 (hex).
Also, in all three formats the last seven (7) bytes of the twenty (20) byte address comprises a six (6) byte end system identifier (or “ESI”) and an one (1) byte selector (or “SEL”). Conventionally, the ESI is an IEEE 802 Media Access Control (or “MAC”) address. Incorporation of the MAC address into the AESA often simplifies the task of mapping AESAs into existing local area networks (or “LANs”). In a typical ATM system, the ESI of an end system is unique for a particular network prefix and is found in the ATM adapter card of the end system. The ESI and the network prefix combine to form a unique nineteen (19) byte address in the network.
In the DCC and ICD formats, the first two (2) bytes following the AFI comprise the initial domain identifier (or “IDI”). The IDI specifies the authority responsible for allocating the subsequent portion of the AESA. In the DCC and ICD formats, the last seventeen (17) bytes is called the domain specific part (or “DSP”) in order to indicate that that portion of the AESA is the portion structured by the authority indicated in the IDI.
The E.164 AESA is based upon the native E.164 format. After the AFI, the next eight (8) bytes comprise a native E.164 address, which is typically an ISDN telephone number. For example, the eight bytes referred to may be comprised of “000014045551212F”. In conventional telecommunications networks using the E.164 AESA, the service provider administers the native E.164 address portion of the E.164 AESA.
The DCC AESA is independent of the native E.164 format. In DCC AESAs, the IDI comprises a two (2) byte data country code. As mentioned above, the DCC AESA and other formats are discussed herein as they relate to, and are used in, the United States, and thus the discussion related to the DCC AESA is ANSI specific. Following the two (2) bytes comprising the IDI is a single-byte DSP format identifier (or “DFI”). The DFI identifies the format of the remainder of the DSP. The three (3) bytes following the DFI comprise the administrative authority (or “AA”) field. The value of the AA field indicates which authority administers the remainder of the DSP, also called the high order domain specific part (or HO-DSP). The HO-DSP is typically structured hierarchically to reflect the network topology or address authorities. An administrative authority may obtain a DCC AESA prefix from, for example, ANSI.
The ICD AESA, like the DCC AESA, is independent of the native E.164 format. The ICD AESA addressing scheme discussed herein is a conventional plan used by BellSouth. In ICD AESAs, the IDI comprises a two (2) byte code for an organization which is responsible for allocating and/or administering the remainder of the AESA. To illustrate a possible layout for the 10 byte HO-DSP, a layout similar to BellSouth's will be described. The first half byte of a HO-DSP of an ICD address comprises reserved, administrative information. The next 1.5 bytes comprise a country code (also called a country field). For example, the country code for the United States is 840. The next byte comprises a region or state code. State codes are listed in FIPS 5-2. For example, the state code for Georgia is 13. The next byte comprises subregion information. In the United States, this field is an encoding of the NPA within a particular state. The next byte comprises the wire center field (also called a switch code). In the United States, this field is a one (1) byte encoding of the wire center within a particular subregion/NPA. This field indicates the wire center containing the switch, which may be the wire center that provides narrowband telephony service to the customer. The next two bytes comprise the termination field. Each customer has an assigned termination field number. The final three bytes comprise the customer part of the HO-DSP. For directly attached customers, this field is set at a value of zero (0). For private network customers, the customer administers this field.
For example, the following string shows a sample encoding of a complete 40 character ICD AESA: 47.0109.0.840.13.02.01.003B.000477.5A29E08443B1.00 (the periods shown in the string are to aid in reading the string only). The AFI field is 47 and the IDI field is 0109. The HO-DSP is 0.840.13.02.01.003B.000477. The administrative portion of the HO-DSP is 0, the country field is 840, which is the code for the United States, and the state code is 13, which is the code for Georgia. The subregion code is 02, the switch code is 01, and the termination field is 003B. Finally, the customer part is 000477.
As can be seen, conventional native E.164 addresses and E.164 AESAs comprise hierarchical, scoped, geographic-based addresses. Although independent of the native E.164 format, conventional DCC and ICD addresses likewise are expected to be administered and deployed as hierarchical, scoped, geographic-based addresses. Conventional addresses are geographically hierarchical in that such addresses contain information about the geographic location of the customer's point of attachment or the customer's switch. Such addresses indicate the physical location of their associated endpoints by multiple-level geographic indicators. Such indicators may be of decreasing geographic scope, i.e., the first field is descriptive of a broader geographical scope, e.g., the United States, than the field following it, e.g., Georgia, which in turn is descriptive of a broader geographical scope, e.g., the 404 area code, than the field following it. The geographic information aids the routing of a call through switches in a communications system to the destination indicated by the address.
Conventional geographic addresses are well-defined hierarchically by geography. For example, in conventional telephone numbers, the first field indicates a country code, the second field indicates a region, generally an area code, the third field indicates a particular exchange, and the fourth field indicates a customer coupled to that exchange. This offers an advantage in that such addresses allow communications systems that are easy to manage. Another advantage is that geographic addresses can be easily summarized so that non-local routing can be based upon a small amount of information. For example, all calls to a non-local NPA can be handed to an interexchange carrier, regardless of the remaining digits in the address.
Geographic-based addresses are contrasted with non-geographic-based addresses. Non-geographic addresses identify the customer to whom the call is to be passed. One example of a non-geographic number is an “800” number, e.g. 1-800-555-1212. An organization-based address is based upon a particular organization rather than geography. A typical Internet IP address is one example of a organization-based address. One advantage of non-geographic addresses is that large customers can connect at multiple locations, and can add to or change these connection points all with a single address prefix.
Both geographic and non-geographic addresses have disadvantages. One disadvantage of geographic addresses is that when a customer moves in physical location, the customer's address must change. Another disadvantage of geographic addresses is that large customers cannot connect at multiple locations, and then add to or change these connection points all with a single address prefix. Geographic addresses are disadvantageous also in that it is highly rigid and generally requires that customers with multiple connections to a system have multiple addresses, each indicating and related to the geographical location of the associated endpoint or connection. Another disadvantage of the conventional geographic address is that because it is hierarchical and geographical in nature, the endpoint having such an address must remain fixed to a particular physical location. One disadvantage of conventional non-geographic addresses is that all switches within the serving area must specifically know that customer's prefix or a central switch must be employed.